kpdvdb 0.5.0

KayPENTAX Disordered Voice Database Reader

This Python module provides functions to interact with KayPENTAX Disordered Voice Database

This module DOES NOT provide the database itself. KayPENTAX Disordered Voice Database is a (discontinued) commercial product, an addon option to KayPENTAX’s Computerized Speech Lab (CSL).

For faster access to the data, copying the database files from the CD-ROM to a local hard drive is highly recommended.

Install

pip install kpdvdb

Use

from kpdvdb import KPDVDB

# to initialize (must call this once in every Python session)
kpdvdb = KPDVDB('<path to CDROM drive or root directory of the database>')

# to list all the data fields
print(kpdvdb.get_fields())

# to list categorical fields' unique values
print(kpdvdb.get_sexes()) # genders
print(kpdvdb.get_locations()) # pathology sites
print(kpdvdb.get_natlangs()) # native languages
print(kpdvdb.get_origins()) # races
print(kpdvdb.get_diagnoses()) # diagnoses

# to get a copy of the full database
df = kpdvdb.query(include_diagnoses=True)

# to get age, gender, diagnoses, and MDVP measures of non-smoking
# subjects with polyp or paralysis, F0 between 100 and 300 Hz
df = kpdvdb.query(["AGE","SEX","DIAGNOSES","MDVP"],
                  DIAGNOSES=["vocal fold polyp","paralysis"],
                  Fo=[100,300],
                  SMOKE=False)

# to get the list of AH NSP files of normal subjects
nspfiles = kpdvdb.get_files('ah',NORM=True)

# to iterate over 'rainbow passage' acoustic data of female pathological subjects
for fs, x, info in kpdvdb.iter_data('rainbow',
                                    auxdata_fields=["AGE","SEX"],
                                    NORM=False, SEX="F"):
  # run the acoustic data through your analysis function, get measurements
  params = my_analysis_function(fs, x)

  # log the measurements along with the age and gender info
  my_logger.log_outcome(*info, *params)

NOTE

Because the database is not public, this module cannot be tested for various platforms via GitHub Action. If you encounter any issues, please post it on GitHub.